It is proposed that prior research has underestimated the strategy capabilities of educable mentally retarded (EMR) children and adolescents. Most cognitive theories of mental retardation have made memory functions a central focus (e.g., Borkowski, Ellis, Spitz, Sternberg, Zeaman) but have treated memory capabilities as an a developmental, static-attribute of cognition and have relied nearly exclusively on verbally-based tasks designed to investigate a single strategy. The proposed research will provide data for a developmental theory of mental retardation that includes situational variables as an aspect of strategy adoption and is based on methodologies that allow multiple strategies, including nonverbal (external) memory strategies. Five experiments will investigate developmental changes in strategy abilities in 11- and 17-year-old EMR children and adolescents and nonretarded mental-age (6 and 9 years) and chronological-age comparison groups. Participants will listen to and follow a series of instructions (e.g., "Put the eraser on the table"). The first experiment will determine if strategy deficits are due, in part, to the use of relatively more difficult tasks with EMR groups. With lower memory loads, participants are expected to be more likely to use external memory strategies (moving objects for the purpose of remembering), indicating that prior studies may have underestimated strategy abilities. The second will use equations derived from the first to predict optimum memory load and will scale situational difference in tasks designed to vary in the degree of situational support for strategy adoption. The magnitude of strategy deficiencies is expected to decrease as situational support increases even though direct instruction on strategy use is not provided. The third experiment will determine if verbal strategy deficiencies are due, in part, to difficulties with "disembedding" a strategy from the response requirements of the task. As the response requirements become more isomorphic with an effective strategy, it is expected that EMR students will demonstrate a wider range of strategy capabilities in an experimenter- absent condition than in an experiment-present condition, implying more cognitive potential than previously observed. In the fifth experiment, EMR students are expected to take advantage of the opportunity to repeat the presentation of a series of instructions an unlimited number of times, demonstrating that they are capable of using task of instructions an unlimited number of times, demonstrating that they are capable of using task resources. These experiments will establish the importance of developmental changes in strategy competence in EMR children and adolescents and the importance of situation contributions to strategy deficiencies.